RankingEffort benchmarks

GPT-5.6 Luna · Max

Compare models

185 measured results · 8 effort settings · 74 benchmark/harness combinations

Model evidence and sources →

Change model

Model

Current configurationSources

Compare with

Choose a second model and its effort to see both profiles and benchmark differences below.

Compare withModel and configuration

Capability profile

One configuration, seven capabilities

Comparison score · 0–100

AGPT-5.6 Luna · Max4 of 7 capabilities supported

Gaps are unknown, not zero. Hollow points are preliminary.

Agentic
A54.2Supported
Hard reasoning
A38.8Supported
Coding
A57.3Supported
Human pref
Unknown
Knowledge
A18.1Supported
Multimodal
A35.3Preliminary
Long context
A66.7Preliminary

Each profile belongs to the named configuration. Capability scores use a shared panel for all axes, distinct from the overall ranking panel; they are not accuracy percentages or direct head-to-head wins.

Capability score evidence

GPT-5.6 Luna · Max: 39.7 overall comparison score. Meets the overall evidence requirements.

6 areas measured. The available task mix affects this estimate; missing areas remain unknown.

Which sources carry the weight?

Vals AI · 13.8% direct comparison weight · independent of the model provider
LiveBench · 13.7% direct comparison weight · independent of the model provider
Cognition · 13.6% direct comparison weight · independent of the model provider
Terminal-Bench · 13.3% direct comparison weight · independent of the model provider
Artificial Analysis · 12.6% direct comparison weight · independent of the model provider
ARC Prize · 12.5% direct comparison weight · independent of the model provider
Epoch AI · 11.8% direct comparison weight · independent of the model provider
Datacurve · 8.6% direct comparison weight · independent of the model provider

Which benchmark families carry the weight?

Terminal-Bench · 15.6% direct comparison weight
FrontierCode · 13.6% direct comparison weight
DeepSWE · 8.6% direct comparison weight
ARC-AGI · 6.8% direct comparison weight
arc-agi-3 · 5.7% direct comparison weight
Other families (34) · 49.7% direct comparison weight

These shares describe direct comparison weight, not fractions of the final score or confidence. Source balancing does not give every family equal weight or remove correlations between overlapping sources.

DeepSWE v1.1 review: Epoch AI reported grading errors in at least 23 of 113 tasks. These published results remain included. Review their limitations in Epoch's review when interpreting this evidence.

How sensitive is this score?

The audit removes one source at a time and refits against the same overall reference panel. These published results use the matching dataset audited 2026-10-20.

Both ranks use the same surviving configurations for each removal. The cohort starts with 444 configurations supported in the original audit, including separate effort settings. These are diagnostic positions, not the default leaderboard ranks. Eligibility is not rechecked after removal.

What changes when a source is removed?
Removed source	Rank before → after	Rank movement	Cohort remaining / lost support	Score after (change)
Artificial Analysis	80 → 81	Down 1	203 remaining / 241 lost	40.1 (+0.3 points)
ARC Prize	88 → 87	Up 1	444 remaining / 0 lost	45.7 (+5.9 points)
Datacurve	88 → 98	Down 10	444 remaining / 0 lost	31.0 (-8.7 points)
LiveCodeBench	88 → 88	No movement	444 remaining / 0 lost	39.7 (-0.0 points)
Scale	88 → 86	Up 2	444 remaining / 0 lost	41.7 (+1.9 points)
Terminal-Bench	88 → 84	Up 4	444 remaining / 0 lost	41.8 (+2.1 points)
Cognition	88 → 92	Down 4	444 remaining / 0 lost	38.3 (-1.4 points)
LiveBench	88 → 84	Up 4	443 remaining / 1 lost	41.7 (+2.0 points)
SWE-rebench	88 → 88	No movement	444 remaining / 0 lost	39.0 (-0.7 points)
Epoch AI	87 → 89	Down 2	431 remaining / 13 lost	40.6 (+0.9 points)
Vals AI	88 → 88	No movement	443 remaining / 1 lost	38.1 (-1.6 points)
Provider reports · Anthropic	88 → 88	No movement	444 remaining / 0 lost	39.7 (+0.0 points)
Provider reports · Moonshot AI	88 → 88	No movement	444 remaining / 0 lost	39.5 (-0.2 points)
Provider reports · Meta	88 → 88	No movement	444 remaining / 0 lost	39.7 (-0.0 points)
Provider reports · Mistral	88 → 88	No movement	444 remaining / 0 lost	39.7 (+0.0 points)
Provider reports · OpenAI	88 → 88	No movement	444 remaining / 0 lost	39.8 (+0.1 points)
All provider reports	88 → 88	No movement	444 remaining / 0 lost	39.5 (-0.2 points)

Lost support means a configuration no longer has a comparison path to the complete reference panel. Its rank is unavailable. A surviving path alone does not establish leaderboard eligibility.

Overlapping sources may remain correlated even with source balancing. Leave-one-source-out movement is a sensitivity diagnostic, not a spread of source-specific ranks or a confidence measure. Scoring methodology · Dated configuration audit

Every benchmark family

Removed evidence	Recomputed score	Change
deepswe	31.0	-8.7
arc-agi-3	41.6	+1.8
arc-agi	41.5	+1.8
simpleqa	41.4	+1.7
terminal-bench	41.3	+1.6
frontiercode	38.3	-1.4
mmlu-pro	41.0	+1.3
livebench-coding	38.4	-1.3
mmmu-pro	40.9	+1.2
frontiermath	38.6	-1.1
hle	40.7	+1.0
livebench-instructions	40.7	+1.0
livebench-language	40.4	+0.7
livebench-data	39.0	-0.7
swe-rebench	39.0	-0.7
aime	39.1	-0.6
vals-excel-modeling	39.2	-0.5
vals-code-migration	39.3	-0.4
livebench-math	40.1	+0.4
gpqa	39.3	-0.4
harvey	40.1	+0.4
vals-finance-agent	39.4	-0.3
swe-atlas-qna	39.4	-0.3
vals-legal-research	40.0	+0.2
vibe-code	39.5	-0.2
terminal-bench-science	39.9	+0.2
omniscience	39.9	+0.2
vals-tax-agent	39.9	+0.2
enterprise-ops	39.9	+0.1
epoch-game-puzzles	39.9	+0.1
ifbench	39.6	-0.1
automationbench	39.6	-0.1
tau-bench	39.6	-0.1
swe-atlas-test-writing	39.6	-0.1
gdp-pdf	39.6	-0.1
gdpval	39.6	-0.1
multi-swe-bench	39.6	-0.1
briefcase	39.6	-0.1
enigma-eval	39.6	-0.1
itbench	39.6	-0.1
livebench-reasoning	39.7	-0.1
aa-lcr	39.7	-0.1
swe-atlas-refactoring	39.7	-0.0
apex-agents	39.7	-0.0
analyst-agent	39.7	+0.0
gmmlu	39.7	+0.0
livecodebench	39.7	-0.0
math500	39.7	+0.0
swe-bench-pro	39.7	-0.0
critpt	39.7	-0.0

What evidence is missing from the fit?

Across all collected settings for this model: 245 observations, 161 contributing, 0 matched without graph weight and 84 excluded. Counts do not establish rank eligibility.

10 · Benchmark family or source protocol has not been reviewed
6 · Original SciCode grading defects are under review; retained as supporting evidence, excluded from aggregate and capability scores. Corrected SciCode-Verified requires a separate protocol review. Sources: https://arxiv.org/abs/2608.04975 and https://artificialanalysis.ai/evaluations/scicode
7 · Evaluator has not been reviewed for independent admission
1 · Effort is not identified
16 · Narrow specialist biological, medical or security evaluation; preserved as raw evidence outside general capability scoring.
16 · Supporting evidence outside the reviewed capability core
23 · No reviewed effort for this catalog observation; any separately collected effort measurement is counted separately
5 · Composite index or publisher-specific internal evaluation lacks a stable general-capability comparison unit.

Inspect every observation and exclusion →

Benchmark results

These are published benchmark scores. Supported effort estimates also appear on the Capability leaderboard. Results from different harnesses occupy different rows. Missing cells stay unknown; no scores are copied between effort levels.

Benchmark / source	Low	Medium	High	XHigh	Max	None	Non-reasoning	Unspecified
AA-Briefcase EloArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Briefcase v1.1 is an agentic knowledge work benchmark developed by Artificial Analysis. AA-Briefcase Elo is a combined metric that aggregates rubric pass rate, analytical quality Elo and presentation Elo · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-10-18.	740.69 Elo Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	934.59 Elo Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	1,170.08 Elo Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	1,267.93 Elo Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	1,342.18 Elo Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	706.04 Elo Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
APEX-Agents-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	35.84% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	Not reported	Not reported
AA-LCR v1.1Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	70% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	75% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	80.33% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	81.67% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	83.67% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	42.67% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
AutomationBench-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of task objectives completed with no guardrail violations · Higher is better · Benchmark developed by Zapier · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	11.69% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	23.07% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	35.59% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	42.6% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	50.21% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	4.97% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
CritPtArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by Argonne and UIUC, with contributions from 60+ researchers globally. CritPt is under review by its authors following external feedback · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	2.57% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	4.86% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	16.57% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	20.57% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	20.57% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	0.29% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
EnterpriseOps-Gym-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Overall task success rate across all eight enterprise domains (oracle tool mode) · Benchmark developed by ServiceNow Research · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	40.82% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	Not reported	Not reported
GDP.pdf: All-pass RateArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of attempts where every atomic criterion passed · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	14.2% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	15.2% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	22.2% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	23.8% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	24% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	6% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
GDPval-AA v2.1 LeaderboardArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Elo rating for performance on real-world work tasks · Anchored to DeepSeek V4.1 Flash (max) at 1600 · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-10-18.	995.33 Elo Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	1,129.07 Elo Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	1,330.1 Elo Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	1,399.92 Elo Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	1,459.2 Elo Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	919.68 Elo Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
GPQA DiamondArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	83.54% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	85.86% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	89.19% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	89.49% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	91.11% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	64.55% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
Humanity's Last ExamArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	19.83% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	25.76% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	33.41% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	36.98% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	39.48% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	7.23% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
ITBench-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Combined public and private Site Reliability Engineering (SRE) tasks · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	40.32% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	Not reported	Not reported
MLCR-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of tasks judged accurate, complete, and concise · Expert + compound (hard) sets · Benchmark developed by Wisedocs · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	6.11% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	8.33% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	12.22% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	16.11% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	19.44% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	Not reported	Not reported
MMMU-ProArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	73.99% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	75.84% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	77.57% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	78.55% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	78.55% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	60.35% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
AA-Omniscience IndexArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct. · Evaluation results measured independently by Artificial Analysis Unit: index. Source collected 2026-10-18.	-14.65 Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	-13.18 Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	-12 Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	-10.77 Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	-10.28 Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	-24.95 Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
AA-Omniscience AccuracyArtificial Analysis · 6000 questions; 1 attempt; correct-answer share; GPT-5.6 Luna medium grader Benchmark details AA-Omniscience Accuracy (higher is better) measures the proportion of correctly answered questions out of all questions, regardless of whether the model chooses to answer · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	39.62% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	40.7% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	41.78% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	42.45% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	42.73% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	28.6% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
SciCodeArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark curated by scientists across 16 disciplines. SciCode is under review following an independent audit of the dataset · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	46.06% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	46.76% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	51.62% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	50.46% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	53.59% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	40.39% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
𝜏³-BankingArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by Sierra Research · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	12.78% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	17.73% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	25.15% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	28.66% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	31.13% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	9.69% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
Terminal-Bench-Science 0.1Artificial Analysis · mini-swe-agent; Terminal-Bench-Science 0.1.0; 70 tasks; pass@1 averaged over 3 repeats Benchmark details Share of tasks passed (pass@1), averaged over three repeats · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	3.33% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	Not reported	Not reported
Terminal-Bench v2.1Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by the Laude Institute and open-source contributors · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	43.45% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	53.18% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	69.66% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	77.9% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	80.9% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	38.95% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
Terminal-Bench v4.0Artificial Analysis · mini-SWE-agent v2.4.6; 66 tasks; pass@1 averaged over 3 repeats; 500 steps; task timeout up to 8 hours Benchmark details Benchmark developed by the Laude Institute, Stanford, and open-source contributors · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	0% Reported settings GPT-5.6 Luna (Low); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-low	0.51% Reported settings GPT-5.6 Luna (Medium); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-medium	2.53% Reported settings GPT-5.6 Luna (High); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-high	3.54% Reported settings GPT-5.6 Luna (Xhigh); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-xhigh	11.62% Reported settings GPT-5.6 Luna (Max); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna	Not reported	1.01% Reported settings GPT-5.6 Luna (Non-reasoning); Model source: https://artificialanalysis.ai/models/gpt-5-6-luna-non-reasoning	Not reported
ARC-AGI-2ARC Prize · v2 Semi-Private Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-10-18.	5.14% Reported settings GPT-5.6 Luna (Low); Source model ID: openai-gpt-5-6-luna-low; 120 tasks	7.36% Reported settings GPT-5.6 Luna (Medium); Source model ID: openai-gpt-5-6-luna-medium; 120 tasks	29.31% Reported settings GPT-5.6 Luna (High); Source model ID: openai-gpt-5-6-luna-high; 120 tasks	47.64% Reported settings GPT-5.6 Luna (XHigh); Source model ID: openai-gpt-5-6-luna-xhigh; 120 tasks	59.54% Reported settings GPT-5.6 Luna (Max); Source model ID: openai-gpt-5-6-luna-max; 120 tasks	Not reported	Not reported	Not reported
DeepSWE v1.1Datacurve · mini-swe-agent Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-10-18.	1.5% Reported settings gpt-5-6-luna; Reported effort: low; 113 unique tasks; 452 attempts; 4 runs	11.3% Reported settings gpt-5-6-luna; Reported effort: medium; 113 unique tasks; 452 attempts; 4 runs	44.2% Reported settings gpt-5-6-luna; Reported effort: high; 113 unique tasks; 452 attempts; 4 runs	56.9% Reported settings gpt-5-6-luna; Reported effort: xhigh; 113 unique tasks; 452 attempts; 4 runs	67.2% Reported settings gpt-5-6-luna; Reported effort: max; 113 unique tasks; 448 attempts; 4 runs	Not reported	Not reported	Not reported
Terminal-Bench 2.1Terminal-Bench · Codex Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	75.73% Reported settings GPT-5.6 Luna; Reported effort: max; 445 trials; recorded 2026-06-26	Not reported	Not reported	Not reported
ARC-AGI-3 · standardARC Prize · Standard Benchmark details ARC-AGI-3; Standard. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-18.	0.17% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-5-6-luna; variant low; public-demo task scores are excluded	0.17% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-5-6-luna; variant medium; public-demo task scores are excluded	0.1% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-5-6-luna; variant high; public-demo task scores are excluded	Not reported	0.18% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group openai-gpt-5-6-luna; variant max; public-demo task scores are excluded	Not reported	Not reported	Not reported
FrontierCode 1.1 · extended-codexCognition · codex Benchmark details FrontierCode 1.1; codex. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-18.	29.02% Reported settings FrontierCode 1.1; extended; 150 tasks; codex; source effort low; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	41.17% Reported settings FrontierCode 1.1; extended; 150 tasks; codex; source effort medium; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	50.52% Reported settings FrontierCode 1.1; extended; 150 tasks; codex; source effort high; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	54.5% Reported settings FrontierCode 1.1; extended; 150 tasks; codex; source effort xhigh; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	55.06% Reported settings FrontierCode 1.1; extended; 150 tasks; codex; source effort max; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0.0013	Not reported	Not reported	Not reported

What these comparisons mean

Each source reports its own effort labels. Matching labels do not establish equal compute budgets or identical fallback behavior. Open Reported settings to inspect the exact model name, agent and task coverage. Multiple results at one level remain visible rather than selecting the best. Source-reported None or Non-reasoning labels are not treated as evidence that an API supports those settings.

These measurements use their own dated source collection. The leaderboard fits identified configurations jointly; it contains no pooled model ratings. Explicit thinking budgets are separate configurations. Read the methodology.

Download all effort benchmark results · 8,915 results across 259 models · Collected 2026-10-18

GPT-5.6 Luna · Max

Compare with

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context

Capability score evidence

Benchmark results